RU2596982C2 - Inspection and determination of location of heart in emission images - Google Patents

Abstract

FIELD: medical equipment.

SUBSTANCE: invention relates to generation of emission images. Method includes steps of: digitizing of emission data; calculating differential emission data; and location of object image corresponding to cyclically variable element in emission data based on differential emission data; calculation and location is performed by means of electronic device, wherein calculation includes step of: calculating difference projection data in space of projection data, containing difference between projection data in different phases of cyclic change, calculation includes steps of: pairwise calculated difference projection data in space of projection data, containing difference between the projection data in different pairs of cyclic change of phases; and combined pairwise difference projection data calculated for different pairs of cyclic change of phases in order to generate difference projection data.

EFFECT: technical result is increasing in accuracy of cardiac shape in emission image.

15 cl, 9 dwg

Description

The following relates to the state of the art of generating imaging images, the state of the art of imaging of the heart, the state of the art of medicine, the state of the art of image processing and the related art.

In the formation of emission images, such as single-photon emission computed tomography (SPECT) or positron emission tomography (PET), a radiopharmaceutical preparation is introduced that includes a radioisotope that emits detectable radiation into an object of examination, such as a therapeutic patient or a sick animal. A radiopharmaceutical is selected or developed to concentrate on the organ or tissue of interest. A radiation sensor located around and / or moving around the test subject measures the detected radiation emitted by the radioisotope, and these measured data are reconstructed to form an emission image or distribution map of the radiopharmaceutical in the test subject. A known application of emission imaging is the imaging of the heart, in which a radiopharmaceutical is selected or developed in order to concentrate in the cardiac organ or in the blood. Therefore, the emission image or map depicts the heart or blood in the cavities (e.g., atria and / or ventricles) of the heart. In some applications, images may be acquired over a period of time to characterize the influx and leaching of the radiopharmaceutical (or blood carrying the radiopharmaceutical) into and out of the heart.

One difficulty regarding the formation of emission images of the heart is that the radiopharmaceutical may also tend to concentrate in other organs, especially the kidneys and liver. This creates many "hot spots" in the emission image, making it difficult to unambiguously establish which object or objects of the image correspond to the heart.

US 2008/0272304 A1 relates to a method and system for automatically identifying and tracking ROI across all planar survey angles of capture of a nuclear imaging system, such as a gamma camera, used to form planar or SPECT images. A temporary change in intensity during the formation of emission projection projection images is measured to identify a region of interest (ROI), such as myocardium. The method automatically tracks the location of the ROI at various planar viewing angles and adjusts the orbit of the sensor head and arranges it to place the ROI inside a predefined preferred area or the so-called "best perception zone" inside the FOV collimator attached to the front side of the surface of the scintillation sensor. After the initial position of the sensor head by the user, the system automatically tracks the location of the ROI while the sensor head (s) are rotated near the patient, and changes the position of the sensor head (s) to suit the ROI inside the optimal collimation region of the FOY sensor.

In “SPATIAL RECONSTRUCTION OF THE OPACIFIED MYOCARDIUM FROM A SMALL NUMBER OF PROJECTIONS” by J Lindenau, O.G.W. Onnasch, J.H. Bursch and P.H. Heintzen on Computers in Cardiology, IEEE Comput. Section 1385, pp. 351-354, examines whether several coronarograms generated by modern dual-projection x-ray equipment provide enough information to estimate the spatial distribution of myocardial blood supply. The modified computed tomography algorithm, the method of projection differences, is described and verified by using the projections of x-ray radiation of the object under study. Coronary angiograms generated through animal experiments serve as “real data”. Using digital subtraction techniques combined with the above CT algorithms, cross sections of the left half of the heart are calculated.

A known approach for locating the heart in an emission image is to use an additional “anatomical” image, such as a transmission computed tomography (CT) image. The anatomical image is spatially recorded using the emission image and then used to determine the location of the heart. This approach has drawbacks in the necessity of accessibility to the second modality of imaging (for example, CT) and the dependence on the exact image registration.

Another known approach is to use a single emission image to determine the location of the heart. Such approaches can be processed manually (for example, an image is displayed through a graphical user interface and the user draws a heart using a mouse pointer or other device) or automatically. Manual processing approaches can be tedious, while existing automated approaches can sometimes mistakenly identify as the heart another organ in which the radiopharmaceutical, such as the liver or kidneys, is concentrated. In both manual and automatic heart location, beats from a beating heart can cause problems, as this introduces a change in time in the data. The set of emission data is usually in the order of minutes to tens of minutes due to safety restrictions on the maximum permissible dose of a radiopharmaceutical and, therefore, low count rates of emission pulses.

One approach to solving this problem is to use “summed” data, which ignores the heartbeat, and the resulting image is blurred. However, the location of the heart can usually be determined in a blurry image. Alternatively, cardiac synchronization may be used based on an electrocardiograph (ECG) or other signal indicating a cardiac cycle. In this case, only the emission data collected in a narrow cardiac phase window is used to restore the image and process the location of the heart. The result is less vague, however, this is due to a significant reduction in emission data for use in the processing of determining the location of the heart and image restoration. (As an example, if the emission data is sampled into eight cardiac phases leading to a coarse phase resolution, only approximately one-eighth of the emission data is used to determine the location of the heart.)

Once the location of the heart is determined, it is useful to check the accuracy of the shape of the heart. Such a check provides control to ensure that neither the liver, nor the kidney, nor any other object is mistakenly identified as a heart, and also preferably provide a more accurate heart shape in the emission image.

The following provides new and improved devices and methods disclosed herein.

In accordance with one aspect disclosed, a method according to claim 1 is provided.

In accordance with another aspect disclosed, a digital processor is configured to perform a method according to claim 1. In accordance with another aspect disclosed, a storage medium is disclosed storing instructions that are executable on a digital processor to execute a method according to claim 1.

In accordance with another aspect disclosed, an apparatus according to claim 13 is provided.

One advantage is the more efficient location of the heart or other cyclically changing element in the emission image.

Another advantage is a more reliable verification of a localized heart or other localized object.

Additionally, the benefits will be apparent to those skilled in the art after reading and understanding the following detailed description.

Figure 1 schematically depicts a SPECT cardiac imaging system, including checking and locating the heart.

Figure 2 schematically depicts the determination of the location of the heart in the emission image, respectively performed by the system according to Figure 1.

Figure 3 schematically depicts the verification and refinement of determining the location of the heart in the emission image, respectively performed by the system according to Figure 1.

Figures 4-6 schematically depict a subprocess for checking whether a heart object with a specific location is essentially circular or oval.

Figures 7-9 schematically depict a subprocess for checking whether a heart object with a specific location is essentially sickle-shaped or notched.

With reference to Figure 1, the imaging system 10 includes an imaging apparatus for acquiring an emission image. An exemplary imaging system 10 is a CardioMD ^™ gamma camera (available from Koninklijke Philips Electronics NV, Eindhoven, The Netherlands) configured to receive SPECT cardiac data from a test subject lying on a support 12 for the test subject using radiation sensors 14 mounted on C-bracket 16. In other embodiments, another type of imaging system may be used to obtain emission data. As another example, the imaging system may be a GEMINI ^™ imaging system (available from Koninklijke Philips Electronics NV, Eindhoven, Netherlands) configured to receive PET imaging data. SPECT, PET, or other emission data are obtained from a sufficiently large range of angular surveys, such as a sweep of 180 ° or 360 °, by moving radiation sensors 14 around the object to be examined and / or by providing a ring or arc of radiation sensors of the required angular amplitude. A suitable radiopharmaceutical is usually administered to the test subject (not shown) by intravenous administration, food intake, inhalation, or by other means of input before obtaining emission data. A time delay can be set between administering the radiopharmaceutical and receiving emission data to allow time for the radiopharmaceutical to accumulate in the heart or other organ of interest. In the illustrative examples in this document, the organ of interest is the heart, but also lungs or another cyclically moving organ or tissue, or an organ or tissue that does not cyclically move, is considered as an organ of interest.

In some embodiments, in which the tissue or organ of interest moves cyclically, an apparatus is provided for tracking (monitoring) cyclic movement. In illustrative examples, cyclic movement is a heartbeat (i.e., cardiac cycle), and an electrocardiograph monitor (ECG) 20 is provided to track the cardiac cycle. Corresponding electrodes or lead wires 22 functionally connect the ECG monitor 20 to the test object to obtain a selected ECG signal (for example, using a standard ECG configuration with 12 lead wires, or a simplified ECG configuration with 4 lead wires, or the like). ) The module 24 for collecting / storing synchronized cardiac emission data collects the emission data obtained by the imaging system 10 and stores the emission data in the sampling intervals of the cardiac phases based on information about the cardiac cycle provided by the ECG monitor 20. For example, in one approach, N is used equal in size to the heart rate sampling intervals, with each heart rate sampling interval having a size T / N in units of seconds, where T is the duration of one heart beat in units of seconds. Uneven size sampling intervals are also considered, for example, to provide a relatively higher temporal resolution for more dynamic parts of the cardiac cycle and a relatively lower temporal resolution for more static parts of the cardiac cycle.

Continuing to refer to Figure 1, the synchronization-based heart locator module 30 processes the synchronized cardiac emission data to determine the location of the heart in the emission data. In some embodiments, the process of determining the location of the heart identifies the heart by automatically identifying emission data that changes over time in accordance with the cardiac cycle, for example, by calculating the differences or derivatives between the emission data at different heart rate sampling intervals. Such differences or derivatives can be calculated in the projection space, wherein the difference or differential projection data is reconstructed by the emission image reconstruction module 32. Alternatively, differences or derivatives can be computed in the image space after reconstructing the emission data of each sampling interval of the cardiac phases by the reconstruction module 32. The optional heart check submodule 34 of the locator module 32 performs a heart check and / or refine the shape of the heart as an additional feature.

Emission data, or a part thereof, identified by the heart locator 30 as emitted from the heart, is reconstructed by the recovery module 32 to generate a heart image (or a series of heart images, for example, in the case of a leaching and / or absorption study of a radiopharmaceutical). In some embodiments, the entire (summed) data set is restored, while in other embodiments, only data from the selected sampling interval or phase sampling intervals is restored. The image or images of the heart are optionally processed by a post-processing system, such as the illustrated cardiac clinical analysis module 38, which performs an analysis to generate quantitative or qualitative clinical information for viewing and / or use by a cardiologist or other medical staff.

Various modules 24, 30, 32, 34, 38 or processing components are respectively embodied by an illustrated computer 40 or other electronic device including a digital processor (e.g., a digital microprocessor, a digital microcontroller, a digital graphics processing unit or the like) and as an option, it also includes analog and / or digital application specific integrated circuit (ASIC) or other hardware. The display device 42 of the computer 40 or some other printing and / or display device, respectively, is used to display emission images or parts thereof and / or images derived from them. A computer 40 or other electronic device also accordingly includes an illustrated keyboard 4 or mouse, trackball, touch pad, or other user input device through which a radiologist or other medical staff controls an imaging system 10 including various modules 24, 30, 32, 34, 38 or processing components.

In addition, various modules 24, 30, 32, 34, 38 or processing components may additionally or alternatively be embodied as an information storage medium storing instructions executed by the illustrated computer 40 or other electronic device to perform the disclosed methods for processing emission data. The storage medium may, by way of illustrative example, include a hard disk drive or other magnetic storage medium, an optical disk or other optical storage medium, random access memory (RAM), read only memory (ROM), flash - memory or another electronic storage medium or the like, or various combinations thereof.

With reference to Figure 1 and with further reference to Figure 2, some relevant synchronization-based approaches for determining the location of the heart, respectively, performed by the module 30 of the heart locator are described. In the beginning, the synchronized projection data 50 of the heart is loaded and sampled onto the cardiac phases in operation 52. For example, operation 52 can be performed by a collection / storage module 24, which in one approach samples the projection data into N equal sizes of the heart rate sampling intervals, each has a sampling interval T / N in units of seconds, where T is the duration of one heart beat in units of seconds. In addition, sampling intervals of unequal sizes are also considered. If unequal sizes of the sampling intervals are used, then in step 52, as an added possibility, the differences in the acquisition time interval are compensated by scaling with the appropriate coefficients for each synchronization sampling interval. In optional operation 54, low-pass filtering frame by frame is used as an additional feature to smooth out projections and reduce interference. The projection data is then differentiated between the synchronization phases in operation 56. In one approach, differential emission data containing differences between the emission data in different phases of a cyclic change are calculated in operation 56. For example, if there are 8 sampling intervals of the cardiac phases, one approach for calculating the differential emission of data should take the difference between the projection data in the intervals 1 and 5 of the sampling of the cardiac phases, and between the projection data in the intervals 2 and 6 of the sampling of the cardiac phases, and dy projection data in intervals 3 and 7 cardiac sampling phase, and between the projection data in intervals 4 and 8 sampling cardiac phases. Projection difference data is computed for each projection viewing angle. In operation 58, the differential projection data from operation 56 is averaged for each projection viewing angle. After the sampling intervals of the cardiac phases are correlated with the cardiac phases, it is expected that the heart is the dominant object in the averaged difference data, while other organs or backgrounds should have a lower signal intensity than the heart. As an additional feature, another noise reduction process may be performed after differentiation operations 56, 58. In the illustrative example of Figure 2, this optional noise reduction is performed as follows. At decision operation 59, a decision is made whether additional noise reduction processing is appropriate (i.e., it is decided whether the difference data is too noisy). If so, then operation 60 of additional low-pass filtering or other noise reduction is performed.

In order to determine the location of the heart in the image space, the combined difference data is restored in step 62, a threshold value is determined for the resulting combined difference image, or else it is processed in step 64 to isolate the heart in the image, and the location is determined in step 66 for the heart. In one appropriate approach, in the reconstruction operation 62, a quick recovery of the averaged difference data to the volumetric image area is performed without any correction (although the application of the correction is also considered). In operation 64, a typical threshold value is expected to be approximately 35% of the maximum value, although the optimal threshold value may vary depending on imaging conditions, signal to noise ratio (SNR), and other factors. In operation 64, the background signal is removed so that the heart is more clearly defined. The positioning step 66 in one approach includes determining the location of the heart in the volumetric image region by calculating the centroid of the array in the image after processing 64. In the positioning operation 66, computing the array centroids in the local area can also be re-applied to search for a finer settings. Other location approaches may also be used in positioning operation 66, such as area extension or segmentation approaches. If it is required to determine the location of the heart in the projection space or region, then in step 68, the location of the heart identified in the image space is accordingly projected back into the projection region to obtain the location of the heart for each frame.

The illustrative processing sequence of Figure 2 performs differentiation 56, 58 in the projection space and then restores the difference projection data in operation 62 to determine the location of the heart in the image space. Alternatively, projection data for each sampling interval of the cardiac phases can be first reconstructed to generate a set of images for different cardiac phases, and differentiation can be performed in the image space (for example, in example 8 sampling intervals of the cardiac phases, calculating: the differences between two images, reconstructed from intervals 1 and 5 of the discretization of the cardiac phases, and the difference between two images reconstructed from intervals 2 and 6 of the discretization of the cardiac phases, and the difference between two images reconstructed from intervals 3 and 7 of the sampling of the heart phases, and the difference between two images reconstructed from intervals 4 and 8 of the sampling of the heart phases, and then combining the calculated difference images similarly to operation 58. In another case, the combining may use averaging, summing, or another merge operation.

With reference to Figure 1 and with further reference to Figures 3-9, some relevant heart test approaches are described respectively performed by the heart test sub-module 34. A heart test functions on a projection image generated from emission data. Figure 3 schematically depicts a complete illustrative process of checking the heart. Not shown in Figure 3 is an optional interference filtering performed as an additional opportunity before entering the heart shape detection routine schematically depicted in Figure 3. The verification processing according to Figure 3 respectively begins with a (suggested) localized heart 70, for example, a derived location procedure hearts described in this document with reference to Figure 2. In operation 72 masking (proposed) localized heart is isolated for processing checks, and in step 74, an initial threshold value 76 is calculated, which is used in the threshold value determination operation 78 to generate a binary mask or representation of the binary image of the (proposed) localized heart.

The heart verification process according to Figure 3 uses a form-based verification process that is based on the expected shape and size of the heart. For example, the heart usually appears as an open circle or partially open circle in emission images (for example, PET or SPECT images). In view of this, the heart check process according to Figure 3 includes a first level shape analysis step 80 in which the (proposed) localized heart is compared with an open circle configuration, and a second level shape analysis step 82 in which the (suggested) localized heart is partially compared with the configuration open circle.

If any analysis of the form 80, 82 is performed, the processing proceeds to the optional adjustment operation 84, in which the (proposed, and now verified) localized heart is corrected based on the results of the verification. An optional adjustment entails providing a more accurate outline of the location of the heart based on a shape analysis.

If both the first analysis of the 80 forms and the second analysis of the 82 forms are unsuccessful, this does not necessarily (at least during the first pass) mean that the proposed localized heart has led to an unsuccessful check. Rather, it could mean that the threshold value of 76 used in form validation analyzes should be adjusted. To this end, in operation 86, the threshold value is increased by the selected amount (for example, 5% in the illustrative Figure 3). In operation 88, it is determined whether the updated threshold value is in a suitable range (for example, less than 100 in the illustrative example), and if so, the process returns to threshold determination operation 78 to repeat the determination of the threshold value using the updated threshold value and repeat analyzes 80, 82 forms.

If it is indicated in operation 88 that the entire suitable range of threshold values has been used and the analyzes 80, 82 of the form were unsuccessful for all threshold values, the process moves to operation 90, in which it is determined whether any other area of the image is the proposed localized heart. If so, then in operation 92, the next proposed localized heart is selected as the proposed localized heart 70 and the process is repeated. On the other hand, if it is determined in step 90 that there is no image area left for testing for the proposed localized heart, then the process proceeds to step 94 of the heart failure test, which reports the failure of the heart test and / or another remedy.

Continuing to refer to Figure 3 and further referring to Figures 4-6, some appropriate approaches are described for performing the first level form analysis operation 80, in which the (proposed) localized heart is compared with an open circle configuration. First, consider Figure 5, which schematically shows the configuration of a typical object in an image. The left side of Figure 5 schematically shows a typical input projection on which the heart is displayed as an open circle H _F and is obscured by secondary objects such as the liver object L. In the upper right of Figure 5, a (proposed) localized heart is shown so that it can be identified by an illustrative processing example according to Figure 2. The lower right of Figure 5 shows the binary mask created in the threshold value determination operation 78 (see Figure 3), and processing by operation 80 of the analysis of the form of the first level, including noise filtering and inversion of the binary mask. Simply connected solid round object H _FV, shown at the bottom right in Figure 5, is a proven heart.

In particular, with reference to Figure 4, the corresponding processing performed in the first level form analysis operation 80 is described. Processing is performed on the binary mask 100 of the localized heart output in step 78 of determining the threshold value (see Figure 3). Interference filtering step 102 is performed to remove smaller components that are too small to be a heart. In an appropriate approach, the interference filtering operation 102 uses a fully connected image analysis (CCA) of the connected components in 2D. Based on the objects remaining after the CCA, in operation 104, the minimum circle size to be detected is calculated. In operation 106, the inverse of the binary mask is performed. In this operation, “on” or black pixels are switched to “off” or white pixels, and vice versa. This transforms shape analysis to detect a simply connected filled circular object. An additional interference filtering operation 108 is performed as an additional feature to remove background noise from the inverted binary mask. In operation 110, in the inverted (and optionally filtered out noise) binary mask, a continuous circular object of suitable size is searched. At operation 112, the success or failure of the operation 80 of the analysis of the form of the first level. If success is reported, then also, as an added opportunity, the best suitable circular-filled object H _{FV is} reported for use in optional adjustment operation 84 (see Figure 3). In one approach, the size of the circle is derived from the number of non-zero pixels in the binary mask.

Figure 6 is an illustration of a successful example of the first level form analysis operation 80 (i.e., an example in which the proposed localized heart is verified). The left side of Figure 6 shows the proposed localized heart in a grayscale image (that is, before any determination of the threshold value). On the right side of FIG. 6, a binary mask is derived from the first level form analysis operation 80.

Continuing to refer to Figure 3 and further referring to Figures 7-9, some appropriate approaches are described for performing a second level form analysis operation 82, in which the (proposed) localized heart is compared with an open partial circle configuration. First, consider Figure 8, which schematically shows the configuration of a typical object in an image. The left side of Figure 8 schematically shows a typical input projection in which the heart is displayed as a partially open circle H _P. In the upper right of Figure 8, a (proposed) localized heart is shown so that it can be identified by an illustrative example of the processing of Figure 2. The lower right in Figure 8 shows the binary mask created in the threshold determination operation 78 (see Figure 3), and processing through the operation 82 of the analysis of the form of the second level, which includes filling the inside of the partially open circle H _P. The fill generates a simply connected solid partially round object H _PV (shown at the very bottom right of Figure 8), which is a verified heart.

In particular, referring to Figure 7, the corresponding processing performed in the second level form analysis operation 82 is described. Processing is again applied over the binary mask 100 of the localized heart output in step 78 of determining the threshold value (see Figure 3). The interference filtering operation 102 is again performed to remove smaller components that are too small to be a heart. In an appropriate approach, the interference filtering operation 102 uses fully-connected image analysis (CCA) of the connected components in 2D, and the CCA of the connected surface is used to detect interference. To improve efficiency, the interference filtering operation 102 may, as an additional feature, be performed only once, and the output is stored and used in the operations 80, 82 of the first and second form analysis.

The interference check is performed in operation 120. If interference is present (even after the CCA), then there will be a significantly larger number of components connected to the surface compared to a fully connected component. If the interference remains after operation 102, then the number of areas on the heart checklist (round cavity or partially round cavity) that are on the final list should not be significantly higher than the size of the binary mask. If it is interference, then this number will be relatively higher than the normal cardiac region, and this information is used to remove the interference.

Based on the objects remaining after the CCA, in operation 122, the minimum size of the partial circle to be detected is calculated. In operation 124, heart filling is performed, as shown schematically in Figure 8. In operation 124 of filling the heart, the remaining part of the partially open circle is filled to produce a simply connected segmented region H _PV corresponding to the part (for example, three quarters) of the filled circle, as shown at bottom right in Figure 8. An additional interference filtering operation 126 is performed as an additional feature to remove background interference after filling 124 hearts. In operation 126, the verified heart is identified based on a partially filled circular object H _PV . At operation 130, the success or failure of the operation 82 of the analysis of the form of the second level. If success is reported, then also as an added opportunity, the best suitable partial filled circular object H _{FV is} reported for use in optional adjustment operation 84 (see Figure 3).

Figure 9 is an illustration of a successful example of the second-level form analysis operation 80 (i.e., an example in which the proposed localized heart is verified). The left side of Figure 9 shows the proposed localized heart in a grayscale image (that is, before any determination of the threshold value). On the right side of Figure 9 there is shown a binary mask derived from the second level form analysis operation 82.

The verification process is described in Figures 3-9 with reference to verification of the proposed localized heart, said circular objects in a more general case may be oval objects. Although the verification process is described in Figures 3-9 with reference to verification of the proposed localized heart, it will be understood that the process can also be used to verify other objects of interest in the emission image, such as the left or right lung.

The positioning approach described for the heart with reference to Figure 2 is applicable to locating another cyclically changing element in the emission data. For example, a cyclically changing element may be a large blood vessel that does not physically move, but which has a cyclically changing emission output signal corresponding to blood flow through a blood vessel.

This application describes one or more preferred embodiments. After reading and understanding the previous detailed description, specialists in the prior art can develop modifications and changes. It is intended that this application be construed as including all such modifications and changes as they fall within the scope of the appended claims or their equivalents.

Claims (15)

1. A method for forming emission images, comprising stages in which:
discretize projection emission data obtained from a cyclically changing element into many different phases of a cyclic change;
calculating the differential emission projection data containing the differences between the emission projection data in different phases of the cyclic change; and
determining the location of the image object corresponding to the cyclically changing element in the emission projection data based on the difference emission projection data;
at the same time, at least the calculation and positioning is carried out by means of an electronic device (40),
wherein the calculation comprises the step of:
calculating the difference projection emission data in the projection data space containing the differences between the emission projection data in different phases of the cyclic change,
characterized in that
the calculation contains the steps in which:
calculating pairwise difference projection emission data in a projection data space containing differences between emission projection data in different pairs of phases of a cyclic change; and
combine the pairwise difference projection emission data calculated for the different pairs of cyclic change phases to generate the difference projection emission data. 2. The method according to claim 1, in which the cyclically changing element is a beating heart, the cyclic change is a heartbeat, and many different phases are many different cardiac phases of a heartbeat. 3. The method according to claim 1, in which the calculation comprises the step of:
calculating the difference projection data in the space of the projection data containing the differences between the projection data in different phases of the cyclic change. 4. The method according to claim 1, in which the location includes stages in which:
recovering said difference projection data in order to generate an image of a locator; and
determine the location of the cyclically changing element in the image of the locator. 5. The method according to claim 4, in which the location further comprises the steps of:
reverse projecting the image of the locator to generate projection data of the locator; and
determine the location of the image object corresponding to the cyclically changing element in the projection data of the locator. 6. The method according to claim 1, additionally containing phase, in which:
reduce interference in said emission data received from a cyclically changing element by low-pass filtering said emission data received from a cyclically changing element. 7. The method according to claim 1, additionally containing a stage in which:
checking the image object with a specific location corresponding to the cyclically changing element, through the verification process, which includes the steps in which:
determining a threshold value of the projection image generated from said emission data in order to generate a mask image,
identify in the mask image one of (i) a hollow circular object, (ii) a hollow oval object, (iii) an object with a circular cavity and (iv) an object with an oval cavity, and
checking the image object at a specific location based on whether the identification operation is successful. 8. The method according to claim 7, in which the operation of identifying the verification process includes the steps of:
invert the image of the mask; and
in the mask image, one of (i) a filled circular object, (ii) a filled oval object, (iii) a partially filled circular object, and (iv) a partially filled oval object is identified. 9. The method of claim 7, wherein the verification process includes a plurality of iterations of operations for determining a threshold value and identification with an increased threshold value used in determining the threshold value in each subsequent iteration. 10. The method according to claim 7, in which the verification process is performed on the area of the projection image, which (i) is the smallest projection image and (ii) is defined by the image object with a specific location corresponding to a cyclically changing element. 11. Digital processor (40), configured to perform the method according to any one of claims 1 to 10. 12. A storage medium that stores instructions that are executable on a digital processor (40) for performing the method according to any one of claims 1 to 10. 13. A device for forming emission images, comprising:
an imaging system (10) configured to receive emission projection data from a cyclically changing element;
instrumentation (20, 22) configured to measure the cyclic change of a cyclically changing element; and
an electronic device (40) configured to determine the location of the image object corresponding to the cyclically changing element in the obtained emission projection data based on the correlation of the time variation of said emission data with the cyclic variation of the cyclically changing element measured by a measuring instrument,
wherein the electronic device (40) is configured to calculate the differential emission projection data in the space of the projection data containing the differences between the emission projection data obtained at different time intervals of the cyclic change of the cyclically changing element measured by the measuring instrument, and determining the location of the image object corresponding to a cyclically changing element, based on differential emission projection data ny
characterized in that
The calculation contains:
calculating pairwise difference projection emission data in a projection data space containing differences between emission projection data in different pairs of phases of a cyclic change; and
combining pairwise difference projection emission data calculated for different pairs of cyclic change phases to generate difference projection emission data. 14. The device according to item 13, in which the cyclically changing element is a beating heart, and the control instrument (20, 22) contains:
an ECG device (20) configured to measure an electrocardiographic signal. 15. The device according to any one of paragraphs.13-14, in which the electronic device (40) correlates the time variation of said emission data in the projection space with the cyclic change of the cyclically changing element measured by a measuring instrument (20, 22).

Images